Method for detecting a synthetic glue nonwoven fabric stain based on optical information

Abstract

The present application relates to the technical field of material defect detection and analysis, and particularly relates to a synthetic glue non-woven fabric stain detection method based on optical information. The method obtains a first image of the non-woven fabric adhesive tape under a white light source and a second image of the non-woven fabric adhesive tape under an alternative light source, calculates the difference degree between the first image and the second image, and when the difference degree is greater than a preset threshold, the non-woven fabric adhesive tape is a suspected defective adhesive tape. Multiple second images of the suspected defective adhesive tape are taken by adjusting the color of the alternative light source, the difference degree between each second image and the first image is calculated, the second image with a difference degree greater than the preset threshold is a contrast image, and the first image and the contrast image are subtracted to obtain a difference image. The difference image is divided into multiple sub-regions, clustering is performed on all the sub-regions to obtain multiple clustering clusters, and the stain area of the suspected defective adhesive tape is obtained according to the best clustering effect. The accuracy of the stain area analysis is improved.

Description

Technical Field

[0001] This invention relates to the field of material defect detection and analysis technology, specifically to a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information. Background Technology

[0002] Medical nonwoven tape is an important medical resource in the treatment and bandaging process. Due to the inherent adhesive properties of nonwoven tape, stains may form on the tape during the production process. Therefore, surface stain detection has always been an important part of quality control during the production of nonwoven tape.

[0003] Existing methods for detecting stains on the surface of nonwoven fabrics are often manual, which requires a lot of manpower and is inefficient. Visual stain detection is often inaccurate due to the texture of the nonwoven fabric surface. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information. This method includes the following steps:

[0005] A first image of the nonwoven tape under a white light source and a second image of the nonwoven tape under a candidate light source are obtained. The difference between the first image and the second image is calculated. When the difference is greater than a preset threshold, the nonwoven tape is a suspected defective tape.

[0006] Multiple second images of the suspected defective tape are captured by adjusting the color of the alternative light source. The difference between each second image and the first image is calculated. The second image with a difference greater than a preset threshold is the comparison image. The difference between the first image and the comparison image is obtained by subtracting the first image from the comparison image.

[0007] The difference image is divided into multiple sub-regions, and all the sub-regions are clustered to obtain multiple clusters. The stain area of ​​the suspected defective tape is obtained based on the best clustering effect.

[0008] The evaluation method for the clustering effect is as follows:

[0009] The first evaluation index is obtained based on the intra-cluster minimum principle of clusters, and the second evaluation index is obtained based on the inter-cluster maximum principle of all clusters.

[0010] The cluster with the most elements among all clusters is selected as the benchmark cluster, and a third evaluation index is obtained based on the difference between the gray-level gradient direction of each pixel in the benchmark cluster and the expected gradient direction.

[0011] All clusters are divided into normal groups and abnormal groups, and a fourth evaluation index is obtained based on the differences between the corresponding image features of the normal groups and the abnormal groups.

[0012] The evaluation index of clustering effect is obtained based on the first evaluation index, the second evaluation index, the third evaluation index and the fourth evaluation index. The optimal clustering effect is when the evaluation index is the largest.

[0013] Preferably, the step of calculating the difference between the first image and the second image includes:

[0014] The grayscale median values ​​of the first image and the second image are obtained respectively, and the difference between the first image and the second image is obtained based on the grayscale median values.

[0015] Preferably, the step of obtaining the difference between the first image and the second image based on the grayscale median value includes:

[0016] The overlapping gray level range of the first and second images is obtained based on the median gray level values ​​of the first and second images. The gray level ranges of the first and second images are updated based on the overlapping gray level range. The probability of each gray level in the updated gray level range of the first image and the probability of each gray level in the updated gray level range of the second image are obtained. The difference between the first and second images is obtained based on the difference in probabilities between corresponding gray levels.

[0017] Preferably, the difference between the first image and the comparison image is the difference between the gray value of each pixel in the first image and the gray value of the corresponding pixel in the comparison image.

[0018] Preferably, the method for dividing the difference image into multiple sub-regions is a region growing algorithm.

[0019] Preferably, the method for obtaining the first evaluation index includes:

[0020] Obtain the average gray value and center coordinates of each sub-region, and calculate the first evaluation index based on the average gray value and the center coordinates:

[0021]

[0022] in, This indicates the primary evaluation indicator; Indicates the first The average gray value of all sub-regions within a cluster; Indicates the first The center coordinates of each cluster; This indicates the number of elements within the cluster. The average gray value of each sub-region; This indicates the first [number] node in the cluster. The center coordinates of each sub-region; This indicates the number of all subregions in the cluster.

[0023] Preferably, the calculation formula for the second evaluation index is as follows:

[0024]

[0025] in, This indicates the second evaluation indicator; This represents the average gray value of all sub-regions within the nth cluster; Indicates the first The center coordinates of each cluster; This represents the average gray value of all clusters. Represents the center coordinates of all clusters; This indicates the number of all clusters.

[0026] Preferably, the step of obtaining the third evaluation index based on the difference between the gradient direction of each pixel in the reference cluster and the desired gradient direction includes:

[0027] The average value of the difference between the gray-level gradient direction and the desired gradient direction of all pixels in the benchmark cluster is taken as the third evaluation index.

[0028] Preferably, the step of obtaining the fourth evaluation index based on the difference between the corresponding image features of the normal group and the abnormal group includes:

[0029] Obtain the gray-level co-occurrence matrix corresponding to the normal group and the gray-level co-occurrence matrix corresponding to the abnormal group, calculate the descriptive operator corresponding to each gray-level co-occurrence matrix, wherein the descriptive operator is contrast and energy; obtain a fourth evaluation index based on the difference between the descriptive operators corresponding to the normal group and the descriptive operators corresponding to the abnormal group.

[0030] Preferably, the evaluation index is negatively correlated with the first evaluation indicator, positively correlated with the second evaluation indicator, negatively correlated with the third evaluation indicator, and positively correlated with the fourth evaluation indicator.

[0031] The present invention has the following beneficial effects: Based on the first and second images of non-woven double-sided adhesive under different point light sources, it is initially determined whether there are stained areas. Further analysis is performed on suspected defective tapes that may have stained areas to obtain difference images. The difference images are then divided into regions and clustered. Stained areas and normal areas are obtained from all the clusters corresponding to the best clustering effect. The evaluation of the clustering effect improves the accuracy of clustering analysis. Based on more accurate clustering results, the detection of stained areas makes the detection results more reliable and accurate. Attached Figure Description

[0032] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a flowchart of a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information, provided as an embodiment of the present invention. Detailed Implementation

[0034] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0036] This application is applicable to the detection of stains on double-sided adhesive tape in synthetic nonwoven fabrics. It analyzes the differences between images under different light sources and obtains the final stain area based on the best clustering effect, thereby improving the accuracy of analysis and identification.

[0037] The following description, in conjunction with the accompanying drawings, details a specific scheme for a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information, provided by the present invention.

[0038] Please see Figure 1 The diagram illustrates a flowchart of a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information, according to an embodiment of the present invention. The method includes the following steps:

[0039] Step S100: Obtain a first image of the nonwoven tape under a white light source and a second image of the nonwoven tape under a candidate light source. Calculate the difference between the first image and the second image. If the difference is greater than a preset threshold, the nonwoven tape is a suspected defective tape.

[0040] Specifically, since non-woven double-sided tape has light-transmitting properties, in this embodiment of the invention, a point light source and a camera are set to acquire images of the non-woven double-sided tape. The point light source is placed on one side of the non-woven double-sided tape, and the camera is placed on the other side. The point light source is white. The first image of the non-woven double-sided tape is obtained by acquiring images of the non-woven double-sided tape using the camera. To increase the accuracy of the analysis, another point light source is set at the same position as the camera. The color of this point light source can be different from white. The positions of the two point light sources and the camera are on the same line. At this time, the second image of the non-woven double-sided tape under this point light source is acquired using the camera.

[0041] Furthermore, since the non-woven double-sided tape itself has certain texture characteristics, it is difficult to detect stains on the surface of the non-woven fabric; however, when stains adhere to the non-woven double-sided tape, they may penetrate, thereby changing the light transmittance of the stained area on the non-woven double-sided tape. Therefore, it is possible to make a preliminary judgment on whether there is a stained area on the non-woven double-sided tape based on the color of the light source.

[0042] When there are no blemishes on the non-woven double-sided adhesive, the outlines of the grayscale histograms corresponding to the first and second images acquired under different light sources should be consistent. However, when blemishes are present, the surface of the non-woven double-sided adhesive changes, resulting in inconsistent light sensitivity. Consequently, the shapes and outlines of the grayscale histograms corresponding to the first and second images under different light sources will change. Therefore, the difference between the first and second images needs to be calculated. The specific method for obtaining the difference between the first and second images is as follows:

[0043] First, the median grayscale value of the first image and the median grayscale value of the second image are obtained respectively, and the difference between the first image and the second image is obtained based on the median grayscale value.

[0044] Obtain the median grayscale value corresponding to the first image; count the number of times each grayscale level appears in the first image to obtain the probability of each grayscale level, and accumulate the probabilities starting from the minimum grayscale level. When the sum of the probabilities equals 0.5, the corresponding grayscale level is recorded as the median grayscale value of the first image, i.e.:

[0045]

[0046] in, This represents the minimum gray level in the first image; Indicates the first image The probability corresponding to each gray level; Indicates the first Each gray level is a grayscale value, i.e., the median grayscale value.

[0047] That is, starting from the probability corresponding to the smallest gray level in the first image, the probabilities are accumulated sequentially in ascending order of gray levels, until the probability is reached at the smallest gray level. When the gray level is n, the sum of the probabilities corresponding to all gray levels is 0.5, therefore the nth gray level... The gray level is the median gray value.

[0048] Based on the method described above for obtaining the same grayscale median value for the first image, the corresponding grayscale median value for the second image is obtained. The corresponding grayscale median values ​​for the first image and the second image are respectively denoted as... , .

[0049] Then, based on the median gray values ​​of the first and second images, the overlapping gray level range of the first and second images is obtained. The gray level ranges of the first and second images are updated based on the overlapping gray level range. The probability of each gray level in the updated gray level range of the first image and the probability of each gray level in the updated gray level range of the second image are obtained. The difference between the first and second images is obtained based on the difference in probabilities between corresponding gray levels.

[0050] Specifically, the overlapping grayscale range between the first and second images is obtained based on the median grayscale value corresponding to each image. Assume the original grayscale range of the first image is... The original grayscale range of the second image is ;in, This represents the smallest gray level in the first image; This represents the maximum gray level in the first image; This represents the smallest gray level in the second image; This represents the maximum gray level in the second image.

[0051] The overlapping grayscale range of the first and second images is then: ;in .

[0052] The gray level range of the first image is updated based on the overlapping gray level range. Update the grayscale range of the second image to... .

[0053] As an example, suppose the grayscale range of the first image is 1000. The grayscale range of the second image is The calculated median grayscale value of the first image is 40, and the median grayscale value of the second image is 50. Therefore, the range of overlapping grayscale levels is... Therefore, the overlapping gray level range updates the gray level range of the first image to... The overlapping gray level range updates the gray level range of the second image to... .

[0054] Finally, the probability of each updated gray level appearing in the first image and the probability of each updated gray level appearing in the second image are obtained, and the difference between the first image and the second image is obtained based on the difference between the probabilities:

[0055]

[0056] in, Indicates the degree of difference; Represents the updated first image The probability corresponding to each gray level; Indicates the updated second image The probability corresponding to each gray level; This represents the median grayscale value corresponding to the first image; This represents the median grayscale value corresponding to the second image; The smallest gray level representing the range of overlapping gray levels; This represents the maximum gray level within the overlapping gray level range.

[0057] It should be noted that in the first image, the first... The gray level and the first gray level in the second image Each gray level is the gray level at the corresponding position within the gray level range; when the difference is greater than the preset threshold, it indicates that there is a large difference between the first image and the second image. Therefore, the non-woven double-sided tape currently being photographed may have stains, and the non-woven double-sided tape is recorded as a suspected defective tape, which needs to be further analyzed.

[0058] Preferably, in this embodiment of the invention, the difference threshold is set to 0.05, that is, when the difference between the first image and the second image is greater than 0.05, further analysis of the nonwoven double-sided adhesive is required.

[0059] Step S200: Adjust the color of the alternative light source to take multiple second images of the suspected defective tape, calculate the difference between each second image and the first image, and the second image with a difference greater than a preset threshold is the comparison image. Subtract the first image from the comparison image to obtain the difference image.

[0060] Specifically, based on step S100, the non-woven double-sided adhesive tape is identified as a potentially defective tape with possible stains. The point light source is transformed based on the difference in material between the stains and the non-woven fabric itself, thereby further identifying the stained area. In this embodiment, a new second image of the potentially defective tape is obtained by adjusting the color of the point light source, denoted as a comparison image. The method for confirming this comparison image is as follows: acquiring new second images of the potentially defective tape under point light sources of different colors, calculating the difference between the second image and the first image in step S100, using the method described in step S100; and determining the color of the point light source corresponding to the second image where the difference is greater than a preset threshold. This second image is then considered the comparison image.

[0061] Preferably, in this embodiment of the invention, the preset threshold is set to 0.1, that is, when the difference between the second image and the first image is greater than 0.1, the color of the point light source is the final color, and the second image at this time is the comparison image.

[0062] It should be noted that when the difference between all the second images and the first images is less than 0.1, the second image with the largest difference among all the differences is selected as the final comparison image.

[0063] Furthermore, a difference image is obtained by subtracting the first image from the comparison image. The difference is calculated by subtracting the gray value of each pixel in the first image from the gray value of the corresponding pixel in the comparison image.

[0064]

[0065] in, Indicates the coordinates in the first image are The grayscale value of the pixel; Indicates the coordinates in the comparison image. The grayscale value of the pixel; This represents the difference between the pixels.

[0066] Similarly, the difference between the first image and the second image at each pixel is obtained, and the image formed by the differences at all pixels is the difference image.

[0067] Step S300: Divide the difference image into multiple sub-regions, cluster all sub-regions to obtain multiple clusters, and obtain the stain area of ​​the suspected defective tape based on the best clustering effect.

[0068] The difference image corresponding to the suspected defective tape is obtained from step S200. The difference image is initially divided into multiple sub-regions. In this embodiment of the invention, the division method adopts a region growing algorithm. A point is randomly selected in the difference image as a seed point, and region growing is performed with the seed point as the center. The growing condition is that the point has the same gray value as the pixel of the seed point. In this way, the difference image is divided into multiple sub-regions.

[0069] Furthermore, the average grayscale value and center coordinates of all pixels within each sub-region are obtained, where the center coordinates are the coordinates of the center point of that sub-region. All sub-regions are clustered based on their average grayscale value and center coordinates, thus grouping areas containing blemishes into the same category, thereby obtaining the blemish areas. In this embodiment, an adaptive k-means clustering algorithm with an adaptive k-value is used. By continuously updating the k-value, the optimal clustering effect is achieved, thereby obtaining the blemish areas in the suspected defective tape. When clustering all sub-regions, the distance between any two sub-regions is:

[0070]

[0071] in, Indicates the distance between two sub-regions; Indicates the first The average gray value of each sub-region; Indicates the first The average gray value of each sub-region; Indicates the first The center coordinates of each sub-region; Indicates the first The center coordinates of each sub-region.

[0072] Preferably, in this embodiment of the invention, the range of k values ​​in the k-means clustering algorithm is set to [1, 20]. By continuously updating the value of k, different clusters are obtained. The best clustering result among multiple clustering operations is selected as the final clustering result.

[0073] The evaluation method for clustering effect is as follows: the first evaluation index is obtained based on the principle of minimum intra-cluster size of clusters, and the second evaluation index is obtained based on the principle of maximum inter-cluster size of clusters; the cluster with the most elements among all clusters is selected as the benchmark cluster, and the third evaluation index is obtained based on the difference between the gradient direction of each pixel in the benchmark cluster and the expected gradient direction; all clusters are divided into normal groups and abnormal groups, and the fourth evaluation index is obtained based on the difference between the corresponding image features of the normal group and the abnormal group; the evaluation index of clustering effect is obtained according to the first evaluation index, the second evaluation index, the third evaluation index and the fourth evaluation index, and the clustering effect is best when the evaluation index is the largest.

[0074] First, the clustering effect should satisfy the condition of minimizing the differences between all sub-regions in each cluster. That is, the first evaluation index of the clustering effect is obtained based on the principle of minimizing intra-cluster differences:

[0075]

[0076] in, This indicates the primary evaluation indicator; Indicates the first The average gray value of all sub-regions within a cluster; Indicates the first The center coordinates of each cluster; This indicates the number of elements within the cluster. The average gray value of each sub-region; This indicates the first [number] node in the cluster. The center coordinates of each sub-region; This indicates the number of all subregions in the cluster.

[0077] Then, the clustering effect should satisfy the condition that the difference between each cluster is maximized. That is, the second evaluation index of the clustering effect is obtained based on the principle of maximizing the difference between clusters:

[0078]

[0079] in, This indicates the second evaluation indicator; This represents the average gray value of all sub-regions within the nth cluster; Indicates the first The center coordinates of each cluster; This represents the average gray value of all clusters. Represents the center coordinates of all clusters; This indicates the number of all clusters.

[0080] Since the stained area on the non-woven double-sided tape only accounts for a small part, the cluster with the most elements in the above clustering should be the normal part of the suspected defective tape. The expected gradient direction of each pixel in the cluster should be related to its own position. Under the illumination of a point light source, the expected gradient direction of each pixel should be a divergent direction centered on the pixel directly opposite the point light source. However, the actual gray-level gradient direction of each pixel in the actual clustering result may differ. The average difference between the gray-level gradient direction and the expected gradient direction of all pixels in the benchmark cluster is taken as the third evaluation index.

[0081] Specifically, the cluster with the most pixels among all clusters is selected as the baseline cluster. The grayscale gradient direction of each pixel in this baseline cluster is then obtained. The calculation method for the gradient direction is a well-known technique and will not be elaborated further. Based on the position of the point light source and the position of the pixel, the expected gradient direction of that pixel can be derived as follows:

[0082]

[0083] in, This indicates the desired gradient direction for that pixel. This indicates the coordinate position corresponding to the pixel. This indicates the coordinates of the pixel directly opposite the point light source, which is the coordinates of the center pixel of the first image.

[0084] The difference between the gray-level gradient direction of a pixel and the desired gradient direction is obtained by calculating the absolute value of the difference between the two directions. Similarly, the differences for all pixels in the baseline cluster are obtained, and a third evaluation metric for the clustering effect is derived based on these differences.

[0085]

[0086] in, This indicates the third evaluation indicator; Indicates the first in the reference cluster The difference between the gray-level gradient direction of each pixel and the expected gradient direction; This indicates the total number of pixels in the baseline cluster.

[0087] Furthermore, considering the significant texture difference between the stained areas and the normal areas of the non-woven synthetic adhesive, all the clusters obtained from clustering are initially classified. The cluster with the most pixels is designated as the normal group, and all other clusters are designated as the abnormal group. The gray-level co-occurrence matrices corresponding to the normal group and the abnormal group are obtained. The descriptive operators corresponding to each gray-level co-occurrence matrix are calculated, and the descriptive operators are contrast and energy. The fourth evaluation index is obtained based on the difference between the descriptive operators corresponding to the normal group and the descriptive operators corresponding to the abnormal group.

[0088] Specifically, the gray-level co-occurrence matrix of the corresponding region image of the normal group and the corresponding region image of the abnormal group are obtained. Based on the gray-level co-occurrence matrix of each group, the corresponding descriptor is obtained. In this embodiment of the invention, the descriptor includes ASM (angular second moment) energy and CON (contrast) contrast. ASM energy is used to reflect the uniformity of image distribution and texture coarseness, and CON contrast is used to reflect the sharpness of the image and the depth of texture grooves. The calculation method of gray-level co-occurrence matrix and its descriptor is a well-known existing technology and will not be described in detail here.

[0089] The fourth evaluation metric for clustering performance, based on the descriptive operators corresponding to the normal and abnormal groups, is:

[0090]

[0091] in, This indicates the fourth evaluation indicator; This represents the contrast of the gray-level co-occurrence matrix corresponding to the normal group; This represents the contrast of the gray-level co-occurrence matrix corresponding to the anomaly group; This represents the energy of the gray-level co-occurrence matrix corresponding to the normal group; This represents the energy of the gray-level co-occurrence matrix corresponding to the anomaly group.

[0092] Based on the first, second, third, and fourth evaluation indicators of the clustering effect obtained above, an evaluation index for the clustering effect is obtained. The evaluation index is negatively correlated with the first evaluation indicator, positively correlated with the second evaluation indicator, negatively correlated with the third evaluation indicator, and positively correlated with the fourth evaluation indicator. Therefore, the evaluation index is:

[0093]

[0094] in, An evaluation index representing the clustering effect; This indicates the primary evaluation indicator; This indicates the second evaluation indicator; This indicates the third evaluation indicator; This indicates the fourth evaluation indicator; This indicates the calculation of the exponential function.

[0095] It should be noted that in the embodiments of this invention, the evaluation indices are all normalized first, second, third, and fourth evaluation indices. Similarly, the evaluation indices corresponding to different k values ​​in the k-means clustering algorithm are obtained. The clustering effect corresponding to the largest evaluation index is the optimal clustering effect. In this case, the cluster with the most pixels among all the clusters obtained is the normal area of ​​the suspected defective tape, and the remaining areas are the stained areas of the suspected defective tape.

[0096] In summary, this invention provides a method for detecting stains on synthetic adhesive nonwoven fabrics based on optical information. By analyzing and subtracting the values ​​of nonwoven double-sided adhesive under different light sources to obtain a difference image, and then performing multiple clustering operations on the difference image, the clustering effect is evaluated each time. The cluster with the most pixels corresponding to the best clustering effect is identified as the normal area of ​​the nonwoven double-sided adhesive, while the areas corresponding to other clusters are identified as the stained areas. This more accurate clustering method improves the accuracy of the analysis by identifying the stained areas.

[0097] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0098] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0099] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An optical information-based synthetic glue nonwoven fabric stain detection method, characterized by, The method comprises the following steps: The method comprises the following steps: Obtain a first image of the non-woven fabric tape under a white light source and a second image of the non-woven fabric tape under an alternative light source, calculate the difference between the first image and the second image, and when the difference is greater than a preset threshold, the non-woven fabric tape is a suspected defective tape; Adjust the color of the alternative light source to take multiple second images of the suspected defective tape, calculate the difference between each of the second images and the first image, and when the difference of the second image is greater than the preset threshold, the second image is a contrast image, and the difference between the first image and the contrast image is obtained to obtain a difference value image; Divide the difference value image into multiple sub-regions, cluster all the sub-regions to obtain multiple clustering clusters, and obtain the stain area of the suspected defective tape according to the best clustering effect; The evaluation method of the clustering effect is: Obtain a first evaluation index based on the minimum principle within the clustering cluster, and obtain a second evaluation index based on the maximum principle between all the clustering clusters; Obtain the cluster with the most elements in all the clustering clusters as a reference cluster, and obtain a third evaluation index based on the difference between the gradient direction of each pixel point in the reference cluster and the expected gradient direction; Divide all the clustering clusters into a normal group and an abnormal group, and obtain a fourth evaluation index based on the difference between the corresponding image features of the normal group and the abnormal group; Obtain an evaluation index of the clustering effect according to the first evaluation index, the second evaluation index, the third evaluation index, and the fourth evaluation index, and the evaluation index is the best clustering effect when it is the largest; The method for obtaining the first evaluation index comprises: in, This indicates the primary evaluation indicator; Indicates the first The average gray value of all sub-regions within a cluster; Indicates the first The center coordinates of each cluster; This indicates the first [number] node in the cluster. The average gray value of each sub-region; This indicates the first [number] node in the cluster. The center coordinates of each sub-region; This indicates the number of all sub-regions in the cluster; Obtain the average gray value and the center coordinates corresponding to each sub-region, and calculate the first evaluation index based on the average gray value and the center coordinates as follows: in, This indicates the second evaluation indicator; This represents the average gray value of all sub-regions within the nth cluster; Indicates the first The center coordinates of each cluster; This represents the average gray value of all clusters. Indicates the center coordinates of all clusters; This indicates the number of all clusters.

2. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 1, characterized by, The calculation formula of the second evaluation index is: The step of calculating the difference between the first image and the second image comprises:

3. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 2, characterized by, Respectively obtain the gray intermediate value of the first image and the gray intermediate value of the second image, and obtain the difference between the first image and the second image based on the gray intermediate value. The step of obtaining the difference between the first image and the second image based on the gray intermediate value comprises:

4. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 1, characterized by, Obtain the overlapping gray level range of the first image and the second image based on the gray intermediate value of the first image and the gray intermediate value of the second image, update the gray level range of the first image and the gray level range of the second image based on the overlapping gray level range, obtain the probability of each gray level in the updated gray level range of the first image, and the probability of each gray level in the updated gray level range of the second image, and obtain the difference between the first image and the second image based on the difference between the probabilities of the corresponding position gray levels.

5. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 1, characterized by, The difference between the first image and the contrast image is the difference between the gray value of each pixel point in the first image and the gray value of the corresponding pixel point in the contrast image.

6. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 1, characterized by, The method for dividing the difference value image into multiple sub-regions is a region growing algorithm. The step of obtaining the third evaluation index based on the difference between the gradient direction of each pixel point in the reference cluster and the expected gradient direction comprises: An average value of a difference between a gray scale gradient direction of all pixel points in the reference cluster and an expected gradient direction is obtained as a third evaluation index.

7. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 1, characterized by, The step of obtaining the fourth evaluation index based on the difference between the image features corresponding to the normal group and the abnormal group comprises: A gray scale co-occurrence matrix corresponding to the normal group and a gray scale co-occurrence matrix corresponding to the abnormal group are obtained, a description operator corresponding to each of the gray scale co-occurrence matrices is calculated, the description operator is contrast and energy, and a fourth evaluation index is obtained based on a difference between the description operators corresponding to the normal group and the abnormal group.

8. The synthetic adhesive nonwoven fabric stain detection method based on optical information according to claim 1, characterized by, The evaluation index is negatively correlated with the first evaluation index, positively correlated with the second evaluation index, negatively correlated with the third evaluation index, and positively correlated with the fourth evaluation index.

Images